Data processing method and corresponding device

The invention discloses a data processing method, which can be applied to a cloud system, an end-cloud cooperative system or an end-side cooperative system, and is particularly suitable for high-performance data analysis (HPDA), such as scenes of artificial intelligence (AI) or image processing and the like. The method comprises the steps that a scheduling node on a cloud side, an end side device or an edge side device can respond to a target computing task of an application, and then data migration cost, task migration cost or computing cost used for running the target computing task is determined according to cost evaluation data and an optimization target of the target computing task. Therefore, the strategy to be executed can be determined according to the corresponding cost, the flexibility of data processing is improved, and the system performance is also improved.